Download V Seminário sobre Riscos, Estabilidade Financeira e Economia

Discussion of Interest Rates and
Default in Unsecured Loan Markets
by Jose Angelo Divino, Edna Souza
Lima and Jaime Orrillo
José Valentim Machado Vicente
Central Bank of Brazil
The Financial Crisis of 2008
Summary and Results
• The authors analyze how interest rates affect the
probability of default.
• First they use a general equilibrium model to this
end.
• They show that there is a positive relationship
between the probability of default and the loan
real interest rate and a negative relationship
between the probability of default and the
economy basic real interest rate.
Summary and Results
• These findings are confirmed by empirical
evidence for the Brazilian economy.
• In the empirical section of the paper the
authors apply survival analysis to estimate the
probability of default under the influence of
macroeconomic conditions.
• The data set consists of loans from a large
Brazilian bank.
Comments
• The idea of the paper is very interesting.
However, the reader would benefit from a paper
revision. First, the exposition of the model and
the results can be improved. For example:
• You do not define theta and gamma in the theoretical
model. I had to read the paper Dubey, Geanakoplos
and Shubik (2005) to understand this point. You
should define these variables in your paper.
• You do not define the function v (remark on page 7).
Comments
– There is a problem in the definition of the payoff
of the risk-free asset. First you assume that the
risk-free asset pays one unit of the commodity in
every state of nature. Then the budget constraint
(Equation 3) must be:
xs  d s  wsh  b  t s rs
Instead of
xs  d s  wsh  (1  r )b  t s rs
Comments
This problem with the payoff of the risk-free asset
remains throughout the text.
• The review of past works can also be improved.
• The authors could compare their results with
other papers that study the macroeconomic
determinants of default probability:
• Carlinga, K., T. Jacobson, K. Roszbach (2007).
Corporate credit risk modeling and the
macroeconomy, Journal of Banking & Finance.
Comments
– Ali, A. and K. Daly (2010). Macroeconomic
determinants of credit risk: Recent evidence from a
cross country study, International Review of Financial
Analysis.
– Chu, V. (2001). Principais fatores macroeconômicos da
inadimplência bancária no brasil. In: BANCO CENTRAL
DO BRASIL. Juros e spread bancário no Brasil:
avaliação de 2 anos do projeto. p. 41-45. Avaible at
www.bcb.gov.br/?spread.
Comments
• The empirical results relate to only one bank.
Do they hold for other banks?
• The paper is divided into two parts:
– Theoretical
– Empirical.
• However, these two parts are quite different:
Comments
Theoretical model
Empirical model
Agents can borrow at the risk free Borrowers have not access to finance
rate
at the risk free rate
Only one risky asset
Many risky assets
Any agent can be a lender or
One agent is a lender and the others
borrower
are borrowers
Default penalties are an increase
Punishment is the inclusion in a
linear function in the size of default
national list of bad payers.
Agents are price takers (perfect
competition)
?
Real assets
Nominal assets
Comments
• Minor points:
– The first relation of Theorem 1 seems to be trivial.
t is a decreasing function of r. Then k is an
increasing function of r (k = 1 – t).
– Is there a bound on sale of assets in the GEI
model?
– Why don’t you extracted real yields from NTN’s
bonds?
Comments
• Minor points:
– There is no description of the tables.
– Why don’t you use lagged macroeconomic
variables to estimate the empirical model?
– The authors should improve the description of
survival analysis.